# Distributed Deep Learning for Precipitation Nowcasting

arXiv: https://arxiv.org/abs/1908.10964 · HTML: https://mlsanity.com/paper/1908.10964/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 28 Aug 2019.

## Built on by (most cited first, 6 of 6)

- [Prediction of severe thunderstorm events with ensemble deep learning and radar data](https://mlsanity.com/paper/2109.09791/) (2021)
- [Statistical Post-Processing for Gridded Temperature Prediction Using Encoder-Decoder-Based Deep Convolutional Neural Networks](https://mlsanity.com/paper/2103.01479/) (2021)
- [Precipitaion Nowcasting using Deep Neural Network](https://mlsanity.com/paper/2203.13263/) (2022)
- [Balancing Computational Efficiency and Forecast Error in Machine Learning-based Time-Series Forecasting: Insights from Live Experiments on Meteorological Nowcasting](https://mlsanity.com/paper/2309.15207/) (2023)
- [Split-n-Chain: Privacy-Preserving Multi-Node Split Learning with Blockchain-Based Auditability](https://mlsanity.com/paper/2503.07570/) (2025)
- [Axial-UNet: A Neural Weather Model for Precipitation Nowcasting](https://mlsanity.com/paper/2504.19408/) (2025)

## Abstract (arXiv)

Effective training of Deep Neural Networks requires massive amounts of data and compute. As a result, longer times are needed to train complex models requiring large datasets, which can severely limit research on model development and the exploitation of all available data. In this paper, this problem is investigated in the context of precipitation nowcasting, a term used to describe highly detailed short-term forecasts of precipitation and other hazardous weather. Convolutional Neural Networks (CNNs) are a powerful class of models that are well-suited for this task; however, the high resolution input weather imagery combined with model complexity required to process this data makes training CNNs to solve this task time consuming. To address this issue, a data-parallel model is implemented where a CNN is replicated across multiple compute nodes and the training batches are distributed across multiple nodes. By leveraging multiple GPUs, we show that the training time for a given nowcasting model architecture can be reduced from 59 hours to just over 1 hour. This will allow for faster iterations for improving CNN architectures and will facilitate future advancement in the area of nowcasting.